Most reactions that build a chiral molecule — one that comes in left- and right-handed versions, chemically identical but mirror images of each other — split close to 50-50 between the two forms. Life does not. Every organism’s proteins are built from the L form of amino acids; its DNA and RNA run on the D form of sugars. The 2026 Nobel Prize in Chemistry, announced on October 7 by the Royal Swedish Academy of Sciences, went to Henri B. Kagan, 95, of Université Paris-Sud, and Kenso Soai, of Tokyo University of Science, “for the discovery of non-linear effects and autocatalysis in asymmetric organic synthesis,” according to the academy’s press release. The two split 12 million Swedish kronor, about $1.2 million.
The property at issue is called chirality. A carbon atom has four bonding sites arranged around a sphere; swap the groups attached to any two of them and you can get a molecule with all the same parts but a different 3D arrangement — its mirror image. Enzymes are binding pockets shaped by evolution to fit one of those arrangements. Feed one the wrong-handed version of its usual substrate and, as Ars Technica’s John Timmer explains, it typically will not bind at all. That is why a world that started chemically racemic — an even mix of both hands — had to resolve itself into the single-handed chemistry every living cell now runs on, and why industrial chemists who want to make a drug that works, rather than a 50-50 mix of a drug and its potentially useless or harmful twin, need a way to force a reaction’s output toward one hand.
A catalyst that plays favorites
Kagan’s contribution addresses the forcing problem. Chemists had assumed that if you build a catalyst that is itself chiral — say, 90 percent one hand and 10 percent the other — the product it churns out would inherit roughly that same 90-10 split. Kagan showed the relationship isn’t linear. When left- and right-handed catalyst molecules pair up with each other (a mismatched, heterochiral pairing) rather than with their own kind, that mismatched pair is markedly less active, sometimes nearly inert. The catalysis gets done almost entirely by the matched (homochiral) pairs, so a small numerical excess in the catalyst mix can produce an outsized excess in the product. GEN’s writeup of the prize dates Kagan’s documentation of this “nonlinear effect” across three separate asymmetric reactions to 1986. That conflicts with the account in the Ars Technica piece drawn on above, which puts three such reactions to “the mid-1960s” — twenty years earlier than every other outlet covering the prize. The Nobel committee’s own dating, and independent corroboration from GEN and from Chemistry World’s live coverage, puts the breakthrough in 1986; Kagan’s underlying career in asymmetric catalysis goes back further, which may account for the mismatch.
Kagan has not published since 2014, per Chemistry World’s reporting. He was, by his own country’s account, already overdue: France’s science minister reportedly complained in 2001 that Kagan had been passed over when that year’s chemistry Nobel went to Barry Sharpless, Ryoji Noyori and William Knowles for separate work on asymmetric catalysis. Twenty-five years is a long wait, even by Nobel standards.
A molecule that makes more of itself
Soai’s work goes after the same imbalance from the other direction: instead of a catalyst that favors one hand, a reaction product that catalyzes more copies of itself. The theoretical case for this was made in 1953 by the physicist Frederick Charles Frank, in a paper titled, with the understatement later singled out by other chemists, “On Spontaneous Asymmetric Synthesis”. Frank’s model required two properties in the same molecule: it had to replicate itself, and it had to suppress replication of its own mirror image (what the paper calls mutual antagonism). Run enough cycles of that and any starting bias, however small, would grow until one hand dominates completely. Frank closed by noting that an experimental demonstration of the idea “may not be impossible.” It took until 2003, fifty years, to find one.
Soai got partway there in 1995, publishing a chiral compound — a 5-pyrimidyl alkanol — that catalyzed its own formation but didn’t purify itself completely. The full version came in 2003: Soai, with Itaru Sato, Hiroki Urabe, Saori Ishiguro and Takanori Shibata, reacted diisopropylzinc with a pyrimidine-5-carbaldehyde and ran the autocatalytic alkoxide product back through the reaction. Starting from an enantiomeric excess of about 0.00005 percent — a bias so small it is barely distinguishable from a true 50-50 mix — three consecutive cycles pushed the excess to over 99.5 percent, according to the paper in Angewandte Chemie International Edition. That is the Soai reaction, and it is the first chemical system shown to do, in a lab, what Frank’s equations said should be possible.
The practical stakes for this kind of control were set by a specific historical failure. In the early 1960s thousands of children were born with birth defects after their mothers took thalidomide, a sedative sold as a racemic mix of both its mirror-image forms; one form was the intended sedative, the other is associated with the teratogenic effects. Kagan began his push to refine asymmetric reactions in the early 1980s partly in response to that episode, aiming to give pharmaceutical chemists a way to manufacture a single, known hand of a drug rather than an uncontrolled mix of two molecules with different effects on the body.
| From | To | How |
|---|---|---|
| Frank's theory, 1953 (autocatalysis plus mutual antagonism) | Soai reaction, 2003 (0.00005% ee to >99.5% ee, 3 cycles) | confirmed, 50 years later |
| Soai: partial autocatalyst, 1995 (5-pyrimidyl alkanol, incomplete purity) | Soai reaction, 2003 (0.00005% ee to >99.5% ee, 3 cycles) | refined |
| Kagan: nonlinear effects, 1986 (mismatched catalyst pairs go inert) | Drug and fragrance synthesis | catalyst design |
| Soai reaction, 2003 (0.00005% ee to >99.5% ee, 3 cycles) | Drug and fragrance synthesis | amplification |
Based on Royal Swedish Academy of Sciences
Henri Kagan and Kenso Soai have provided a solution to a chemical mystery that is over a century old: how homochirality can emerge spontaneously. The chemical reactions they have developed are spectacular.
That line is from Heiner Linke, chair of the Nobel Committee for Chemistry, in the academy’s press release. A committee member, Peter Somfai, described the Soai reaction in plainer terms, as producing “an excess of an enantiomer which then forms copies of itself,” calling it, per GEN’s account, “probably the coolest experiment in organic chemistry.”
Soai was out shopping when the committee reached him by phone. “This is the most exciting day in my life, I am very glad to share this prize with professor Henri Kagan,” he told the press conference, adding, by one account, that he hoped the prize would make people “recognise the importance of the chirality of molecules” and that the study of it “should develop even more and more” for the next generation of chemists. Chemistry World’s live coverage reports he cut the call short with the explanation that his country’s prime minister was trying to reach him.
None of the chemistry honored here was run on an actual biomolecule, or anywhere near the conditions of the early Earth; Kagan’s and Soai’s reactions were built, in a modern lab, specifically to demonstrate that such amplification is chemically possible, not to recreate how it happened four billion years ago. What the natural record offers instead are much smaller biases: amino acids extracted from the Murchison meteorite show an L-enantiomeric excess of isovaline measured at up to 18.5 percent in one fragment, according to a 2009 PNAS study, with other fragments measuring anywhere from near zero to around 20 percent. That is a long way short of the more-than-99-percent the Soai reaction produces, and nobody has shown an autocatalytic amplifier of that kind operating on a molecule that actually appears in biochemistry, rather than a pyrimidine aldehyde selected because it happens to amplify well. The open number is still which molecules, if any, in the chemistry that preceded life had both of Frank’s required properties — self-replication and mutual antagonism toward their own mirror image — at any bias close to the small ones meteorites actually show.